Re: [PATCH v7 07/20] perf record: Introduce data transferred and compressed stats

["Bayduraev, Alexey V" <alexey.v.bayduraev@linux.intel.com>]

On 30.06.2021 15:19, Bayduraev, Alexey V wrote:
> 
> Hi,
> 
> On 30.06.2021 11:05, Riccardo Mancini wrote:
>> Hi,
>>
>> On Tue, 2021-06-22 at 11:42 +0300, Alexey Bayduraev wrote:
>>> Introduce bytes_transferred and bytes_compressed stats so they
>>> would capture statistics for the related data buffer transfers.
>>>
>>> Acked-by: Andi Kleen <ak@linux.intel.com>
>>> Signed-off-by: Alexey Bayduraev <alexey.v.bayduraev@linux.intel.com>
>>> ---
>>>  tools/perf/builtin-record.c | 64 +++++++++++++++++++++++++++++--------
>>>  tools/perf/util/mmap.h      |  3 ++
>>>  2 files changed, 54 insertions(+), 13 deletions(-)
>>>
>>> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
>>> index 38bb5afbb359..c9fd31211600 100644
>>> --- a/tools/perf/builtin-record.c
>>> +++ b/tools/perf/builtin-record.c
>>> @@ -198,6 +198,11 @@ static int record__write(struct record *rec, struct mmap
>>> *map __maybe_unused,
>>>                 return -1;
>>>         }
>>>  
>>> +       if (map && map->file) {
>>> +               map->bytes_written += size;
>>> +               return 0;
>>> +       }
>>> +
>>>         rec->bytes_written += size;
>>>  
>>>         if (record__output_max_size_exceeded(rec) && !done) {
>>
>> This breaks the --max-size option in parallel trace mode.
>> Maybe, we could change record__output_max_size_exceeded to check all
>> thread_data->maps and sum up bytes_written. Concurrency should not be an issue
>> since it's not required to stop at exactly the max-size (this is also the
>> current behaviour).
>> Otherwise, we could atomically increase an accumulator inside struct record
>> (maybe rec->bytes_written could be reused, but I'm not sure if it's used by the
>> main thread in the parallel mode), and check it for exceeded max size.
> 
> Thanks, currently threaded mode doesn't use rec->bytes_written and atomic64_t
> framework is not available in kernel/tools.
> I think better is to calculate sum of map->bytes_written like for threads 
> statistics.

Just a little clarification: threaded mode uses rec->bytes_written for user
records (perf.data/data), whereas maps->bytes_written for per-cpu records
(perf.data/data.CPU).
Thus, the total amount of data transferred is their sum.
 
Regards,
Alexey

> 
> Regards,
> Alexey
> 
>>
>> Thanks,
>> Riccardo
>>
>>
>>> @@ -215,8 +220,8 @@ static int record__write(struct record *rec, struct mmap
>>> *map __maybe_unused,
>>>  
>>>  static int record__aio_enabled(struct record *rec);
>>>  static int record__comp_enabled(struct record *rec);
>>> -static size_t zstd_compress(struct perf_session *session, void *dst, size_t
>>> dst_size,
>>> -                           void *src, size_t src_size);
>>> +static size_t zstd_compress(struct zstd_data *data,
>>> +                           void *dst, size_t dst_size, void *src, size_t
>>> src_size);
>>>  
>>>  #ifdef HAVE_AIO_SUPPORT
>>>  static int record__aio_write(struct aiocb *cblock, int trace_fd,
>>> @@ -350,9 +355,13 @@ static int record__aio_pushfn(struct mmap *map, void *to,
>>> void *buf, size_t size
>>>          */
>>>  
>>>         if (record__comp_enabled(aio->rec)) {
>>> -               size = zstd_compress(aio->rec->session, aio->data + aio->size,
>>> -                                    mmap__mmap_len(map) - aio->size,
>>> +               struct zstd_data *zstd_data = &aio->rec->session->zstd_data;
>>> +
>>> +               aio->rec->session->bytes_transferred += size;
>>> +               size = zstd_compress(zstd_data,
>>> +                                    aio->data + aio->size,
>>> mmap__mmap_len(map) - aio->size,
>>>                                      buf, size);
>>> +               aio->rec->session->bytes_compressed += size;
>>>         } else {
>>>                 memcpy(aio->data + aio->size, buf, size);
>>>         }
>>> @@ -577,8 +586,22 @@ static int record__pushfn(struct mmap *map, void *to,
>>> void *bf, size_t size)
>>>         struct record *rec = to;
>>>  
>>>         if (record__comp_enabled(rec)) {
>>> -               size = zstd_compress(rec->session, map->data,
>>> mmap__mmap_len(map), bf, size);
>>> +               struct zstd_data *zstd_data = &rec->session->zstd_data;
>>> +
>>> +               if (map->file) {
>>> +                       zstd_data = &map->zstd_data;
>>> +                       map->bytes_transferred += size;
>>> +               } else {
>>> +                       rec->session->bytes_transferred += size;
>>> +               }
>>> +
>>> +               size = zstd_compress(zstd_data, map->data,
>>> mmap__mmap_len(map), bf, size);
>>>                 bf   = map->data;
>>> +
>>> +               if (map->file)
>>> +                       map->bytes_compressed += size;
>>> +               else
>>> +                       rec->session->bytes_compressed += size;
>>>         }
>>>  
>>>         thread->samples++;
>>> @@ -1311,18 +1334,15 @@ static size_t process_comp_header(void *record, size_t
>>> increment)
>>>         return size;
>>>  }
>>>  
>>> -static size_t zstd_compress(struct perf_session *session, void *dst, size_t
>>> dst_size,
>>> +static size_t zstd_compress(struct zstd_data *zstd_data, void *dst, size_t
>>> dst_size,
>>>                             void *src, size_t src_size)
>>>  {
>>>         size_t compressed;
>>>         size_t max_record_size = PERF_SAMPLE_MAX_SIZE - sizeof(struct
>>> perf_record_compressed) - 1;
>>>  
>>> -       compressed = zstd_compress_stream_to_records(&session->zstd_data, dst,
>>> dst_size, src, src_size,
>>> +       compressed = zstd_compress_stream_to_records(zstd_data, dst, dst_size,
>>> src, src_size,
>>>                                                      max_record_size,
>>> process_comp_header);
>>>  
>>> -       session->bytes_transferred += src_size;
>>> -       session->bytes_compressed  += compressed;
>>> -
>>>         return compressed;
>>>  }
>>>  
>>> @@ -2006,8 +2026,10 @@ static int record__start_threads(struct record *rec)
>>>  
>>>  static int record__stop_threads(struct record *rec, unsigned long *waking)
>>>  {
>>> -       int t;
>>> +       int t, tm;
>>> +       struct mmap *map, *overwrite_map;
>>>         struct thread_data *thread_data = rec->thread_data;
>>> +       u64 bytes_written = 0, bytes_transferred = 0, bytes_compressed = 0;
>>>  
>>>         for (t = 1; t < rec->nr_threads; t++)
>>>                 record__terminate_thread(&thread_data[t]);
>>> @@ -2015,9 +2037,25 @@ static int record__stop_threads(struct record *rec,
>>> unsigned long *waking)
>>>         for (t = 0; t < rec->nr_threads; t++) {
>>>                 rec->samples += thread_data[t].samples;
>>>                 *waking += thread_data[t].waking;
>>> -               pr_debug("threads[%d]: samples=%lld, wakes=%ld,
>>> trasferred=%ld, compressed=%ld\n",
>>> +               for (tm = 0; tm < thread_data[t].nr_mmaps; tm++) {
>>> +                       if (thread_data[t].maps) {
>>> +                               map = thread_data[t].maps[tm];
>>> +                               bytes_transferred += map->bytes_transferred;
>>> +                               bytes_compressed += map->bytes_compressed;
>>> +                               bytes_written += map->bytes_written;
>>> +                       }
>>> +                       if (thread_data[t].overwrite_maps) {
>>> +                               overwrite_map =
>>> thread_data[t].overwrite_maps[tm];
>>> +                               bytes_transferred += overwrite_map-
>>>> bytes_transferred;
>>> +                               bytes_compressed += overwrite_map-
>>>> bytes_compressed;
>>> +                               bytes_written += overwrite_map->bytes_written;
>>> +                       }
>>> +               }
>>> +               rec->session->bytes_transferred += bytes_transferred;
>>> +               rec->session->bytes_compressed += bytes_compressed;
>>> +               pr_debug("threads[%d]: samples=%lld, wakes=%ld,
>>> trasferred=%ld, compressed=%ld, written=%ld\n",
>>>                          thread_data[t].tid, thread_data[t].samples,
>>> thread_data[t].waking,
>>> -                        rec->session->bytes_transferred, rec->session-
>>>> bytes_compressed);
>>> +                        bytes_transferred, bytes_compressed, bytes_written);
>>>         }
>>>  
>>>         return 0;
>>> diff --git a/tools/perf/util/mmap.h b/tools/perf/util/mmap.h
>>> index c4aed6e89549..c04ca4b5adf5 100644
>>> --- a/tools/perf/util/mmap.h
>>> +++ b/tools/perf/util/mmap.h
>>> @@ -46,6 +46,9 @@ struct mmap {
>>>         int             comp_level;
>>>         struct perf_data_file *file;
>>>         struct zstd_data      zstd_data;
>>> +       u64                   bytes_transferred;
>>> +       u64                   bytes_compressed;
>>> +       u64                   bytes_written;
>>>  };
>>>  
>>>  struct mmap_params {
>>
>>